Langmuir 2004, 20, 1473-1480
1473
Interaction of Asphaltenes with Nonylphenol by Microcalorimetry Daniel Merino-Garcia and Simon I. Andersen* IVC-SEP, Department of Chemical Engineering, Technical University of Denmark, DK-2800 Lyngby, Denmark Received October 8, 2003. In Final Form: December 10, 2003 This paper shows the work performed in the study of the capability of isothermal titration calorimetry (ITC) to characterize the interaction between petroleum asphaltenes with a model molecule, namely, nonylphenol. ITC is widely used in biochemistry to study the interaction of proteins with ligands. The intention is to transfer the knowledge into the asphaltene field, with the aim of getting a better understanding of the mechanism of interaction, as well as the energies involved in this process. Calorimetric experiments show that nonylphenol has a complex mechanism of interaction with asphaltenes in toluene, including more than one process. Several models have been used to fit the experimental data. The enthalpies calculated with a model based on polymerization are in the order of -1 to -7 kJ/mol, which are very close to the hydrogen bond energies. This shows the capability of ITC to provide experimental data to the modeling of asphaltene behavior. The number of sites of interaction has been inferred by means of a model taken from protein-ligand science. The values obtained are in the range two to five sites per molecule, assuming an average Mw of 1000 units.
Introduction Asphaltenes are one of the heaviest fractions of crude oil. They are defined as the fraction of petroleum that is insoluble in n-alkane (i.e., n-heptane) and soluble in toluene. This weak definition allows the presence in this fraction of molecules of a wide range of sizes, heteroatom content, and aromaticity. The importance of asphaltenes is based on the problems they cause in the recovery and refining of petroleum. Together with waxes, they are a big nuisance for field engineers. They have a great tendency to associate, leading to precipitation and facility plugging problems. Asphaltenes are also responsible for the poisoning of catalysts, especially at elevated temperatures and pressures.1 A large bibliography about asphaltenes is available.2 However, many questions are still open about the mechanism of association and the best way of avoiding it. Resins are another petroleum fraction also found in vacuum residua. Compared to asphaltene research, resin studies are scarce.3 The resin fraction contains also heteroatomic groups, together with long nonpolar paraffinic groups.4 They are the transition between the polar asphaltenes and the nonpolar dispersing media. Resins are usually separated from the deasphalted oil by adsorption on surface-active materials, but sometimes they are also defined as the fraction of deasphalted oil that is insoluble in liquid propane.5 The separation procedure has a significant impact on the properties of the resulting resins.6 Resin interaction with asphaltenes is believed to be a combination of van der Waals, charge transfer, Coulombic, and exchange-repulsion forces.7 Resins are supposed to act as stabilizers of asphaltenes, * To whom correspondence may be addressed. E-mail, sia@ kt.dtu.dk. (1) Seki, H. Fuel Process. Technol. 2001, 69 (3), 229. (2) Speight, J. G. The Chemistry and Technology of Petroleum, 3rd ed.; Marcel Dekker Inc.: New York, 1999. (3) Andersen, S. I.; Speight, J. G. Pet. Sci. Technol. 2001, 19 (1&2), 1. (4) Koots, J. A.; Speight, J. Fuel 1975, 54, 179. (5) Schwager, I.; Yen, T. F. Fuel 1978, 57, 100. (6) Goual, L.; Firoozabadi, A. AIChE J. 2002, 48 (11), 2646. (7) Murgich, J. Pet. Sci. Technol. 2002, 20 (9&10), 983.
as the removal of resins causes the precipitation of asphaltenes.4 To remediate asphaltene fouling, the concentration of resin-like material is increased by adding chemicals such as alkyl-phenolic surfactants. These molecules would act as synthetic resins that help in the stabilization of asphaltenes. However, to be able to optimize the inhibitorasphaltene interaction and therefore the performance and cost of operation, knowledge of the interaction between asphaltenes and other oil fractions or additive chemicals is of great importance. The stabilization of asphaltenes with surfactants depends on the polarity of the headgroup;8 this shows again the importance of the polar nature of asphaltenes in their stability and aggregation behavior. For instance, the addition of nonylphenol delays the onset point of the flocculation of asphaltenes.9 Moschopedis et al.10 proved that the addition of phenol decreased the measured molecular weight of asphaltenes, as if they broke the aggregates and dispersed the asphaltene molecules. Chang et al.8 reported that 7 wt % of nonylphenol was able to completely solubilize 1 wt % asphaltenes in n-heptane. Leon et al.11 showed, however, that the capacity of stabilizing asphaltenes is higher for native resins than for amphiphiles, even if they adsorb in lower quantities. The adsorption to asphaltene particles in n-heptane can be attributed to both multilayer adsorption and penetration of resins in the microporous structure of asphaltenes. Chang et al.12 reported by small-angle X-ray scattering that asphaltene-nonylphenol aggregates are slightly bigger than asphaltene aggregates, but resins11 do decrease the size of asphaltene aggregates. This suggests that nonylphenol does not disperse asphaltenes in the same way as native resins do. In this study, nonylphenol has been chosen as a model resin to study its interaction with asphaltenes. These experiments will help in the understanding of the mechanism of stabilization by (8) Chang, C. L.; Fogler, H. S. Langmuir 1994, 10, 1749. (9) Clarke, P. F.; Pruden, B. B. Pet. Sci. Technol. 1998, 16 (3&4), 287. (10) Moschopedis, S. E.; Speight, J. J. Mater. Sci. 1977, 12, 990. (11) Leon, O.; Contreras, E.; Rogel, E.; Dambakli, G.; Acevedo, S.; Carbognani, L.; Espidel, J. Langmuir 2002, 18, 5106. (12) Chang, C. L.; Fogler, H. S. Langmuir 1994, 10, 1758.
10.1021/la035875g CCC: $27.50 © 2004 American Chemical Society Published on Web 01/14/2004
1474
Langmuir, Vol. 20, No. 4, 2004
surfactants. This work will provide as well a solid background that will later be used in the interaction of native resins with asphaltenes. Isothermal titration calorimetry (ITC) is widely used in biochemistry in the study of ligand-protein interactions.13 However, it has never had an extended application to petroleum species.14-16 Several models have been used in the past to model the behavior of asphaltenes both in solution and in live oil from several approaches, including regular solution theory,17 micellization,18,19 and lately applications of the SAFT equation.20-22 Murgich et al.23 and Ortega-Gonzalez et al.24 performed molecular simulations of the interaction between asphaltene and resin model molecules. All these models include numerous parameters that require either estimation or fitting without knowledge of actual ranges or magnitudes. The experimental values provided by a technique such as ITC would be useful in order to decrease the number of estimations and also to check the quality of the model, by comparing the experimental data with the values obtained in the fitting procedure. Materials and Methods Toluene (spectroscopic grade) was obtained from Rathburn. It was dried with molecular sieves to minimize the influence of water, as it has been previously shown to play a significant role in the interaction between asphaltenes and resins.15 Asphaltenes from different sources have been used in this study: LM1, LM2, NM1 (Venezuela), Alaska 95, KU and Yagual (Mexico), Lagrave (France), and Ca30. They have been obtained by following a modified IP143 procedure.25-26 Asphaltenes have been washed with n-heptane to remove the last traces of resins and waxes that could have been coprecipitated with the asphaltenes or just occluded by them. A mixture 70:30 of toluene and acetone was used to fractionate KU asphaltenes. This mixture was chosen in the light of previous fractionation studies.27 For LM1 asphaltenes, the ratio used was 60:40. Although affecting the distribution of species, the intention was to obtain an approximately 50-50 wt % fractionation, and hence the ratio of acetone was changed depending on the origin of the asphaltene. The solution was mixed in an ultrasonic bath for about 60 min. The samples were left overnight, and then centrifuged for 5 min at 3000 rpm and then the insoluble (INS) and soluble (SOL) fractions were separated and the solvent was evaporated.26 CHNS elemental analyses were carried out using standard procedures by DB Lab in Odense, Denmark. (13) Freire, E.; Mayorga, O. L.; Straume, M. Anal. Chem. 1990, 62 (18), 950A. (14) Andersen, S. I.; Birdi, K. S. J. Colloid Interface Sci. 1991, 142 (2), 497. (15) Andersen, S. I.; del Rio, J. M.; Khvostitchenko, D.; Shakir, S.; Lira-Galeana, C. Langmuir 2001, 17, 307. (16) Murgich, J.; Lira-Galeana, C.; Merino-Garcia, D.; Andersen, S. I., del Rio, J. M. Langmuir 2002, 18, 9080. (17) Wang, J. X.; Buckley, J. S. Energy Fuels 2001, 15 (5), 1004. (18) Pan, H.; Firoozabadi, A. AICHE J. 2000, 46 (2), 416. (19) Victorov, A. I.; Smirnova, N. A. Fluid Phase Equilib. 1999, 158160, 471. (20) Wu, J.; Prausnitz, J. M.; Firoozabadi, A. AIChE J. 2000, 46 (1), 197. (21) Buenrostro-Gonzalez, E.; Gil-Villegas, A.; Wu, J.; Lira-Galeana, C. Proceedings 2002 International Conference on Heavy Organic Deposition; Lira-Galeana, C., Ed.; Mexico, 2002. (22) Ting, D.; Hirasaki, G. J.; Chapman, W. G. Pet. Sci. Technol. 2003, 21 (3&4), 647. (23) Murgich, J.; Strausz, O. P. Pet. Sci. Technol. 2001, 19 (1&2), 231. (24) Ortega-Gonzalez, A.; Lira. Galeana, C.; Ruiz-Morales, Y.; Cruz, S. A. Pet. Sci. Technol. 2001, 19 (1&2), 245. (25) IP 143/90. Standards for Petroleum and its Products. Institute of Petroleum, London: 1985. (26) Merino-Garcia, D.; Andersen, S. I. Pet. Sci. Technol. 2003, 21 (3&4), 507. (27) Buenrostro-Gonzalez, E.; Andersen, S. I.; Garcia-Martinez, J. A.; Lira-Galeana, C. Energy Fuels 2002, 16 (3), 732.
Merino-Garcia and Andersen FT-IR Spectroscopy. The equipment used is a Perkin-Elmer PARAGON 1000 FT-IR spectrometer with a Specac cell of 0.5 mm path length and NaCl windows. The interest has been focused on the hydroxyl stretching region (3700-3200 cm-1). Isothermal Titration Calorimetry (ITC). The experimental setup of the calorimeter (VP-ITC 2000 Microcal) has been previously described.26 All tests are carried out at 303.0 ( 0.1 K. The calorimeter consists of two cells: a reference cell filled with pure solvent, and a sample cell that is used as a titration cell. It initially contains a solution of asphaltenes in toluene, and the reagent (nonylphenol or resins) is added into it by means of a syringe in small steps of 5.0-10.0 µL. The isothermal conditions are kept by means of a control system that supplies or takes heat, depending on the heat developed in the interaction among the several species. Several concentrations have been used, both in the syringe (22.7, 90.9, 181.8, 299.8, and 469.7 mM of nonylphenol, approximately 5, 20, 40, 66, and 100 g/L) and in the cell (0.25, 1.0, and 10.00 g/L of asphaltenes). There are several processes going on in the cell in this type of experiment: The injected toluene causes a decrease in asphaltene concentration that leads to a rearrangement of the equilibrium between free and aggregated asphaltenes. The solution of nonylphenol in the syringe is also in equilibrium between free and hydrogen-bonded molecules. The injection of this solution into a greater volume causes as well the dissociation of nonylphenol aggregates. The injection process also leads to some frictional heat. Finally, nonylphenol interacts with the asphaltenes, theoretically by means of the formation of hydrogen bonds. Several models have been tested in order to fit the experimental data to an expression that allows the calculation of the thermodynamic properties of the interaction of asphaltenes with the ligands. In the first approach (model 1), the titration is modeled with three equations, and each of them has two parameters: the equilibrium constant (K) and heat of interaction (∆H).
Interaction NP-ASP: A2 + N w A2N {K, ∆H}
(1)
Breaking of ASP aggregates: A2 w A + A {KA, ∆HA}
(2)
Breaking of NP-NP bonds: N2 w N + N {KN, ∆HN}
(3)
This approach is based on the chemical theory,28 and it considers that the only heat developed is due to the formation of new species.26 Notice that nonylphenol is assumed to react with asphaltene aggregates (A2). A model in which NP reacts with the monomer A was as well proposed with very poor results. After each injection, the concentration of the different compounds changes and a new equilibrium has to be reached. The differences in concentrations between the moment after the injection and the equilibrium allow calculation of the number of physical intermolecular bonds formed and broken in the reorganization process. The total heat developed q in each injection is
q ) nA-N (mol)*(∆H) + nA-A (mol)*(-∆HA) + nN-N (mol)*(-∆HN) (4) nN-A is the number of ASP-NP bonds formed and nA-A and nN-N are the number of ASP-ASP and NP-NP bonds broken, respectively. The parameters of the models are optimized so that the heat calculated with eq 4 is as close as possible to the experimental heat. Six parameters are too many to be optimized in only one run. It is necessary to perform reference tests to calculate some parameters. The values of {KA, ∆HA} are taken from the titrations in which asphaltenes are injected into pure dried toluene.16 To (28) Prausnitz, J. M.; Lichtenthaler, R. N.; Azevedo, E. G. Molecular Thermodynamics of Fluid-Phase Equilibria; Prentice Hall: Englewood Cliffs, NJ, 1999.
Interaction of Asphaltenes with Nonylphenol
Langmuir, Vol. 20, No. 4, 2004 1475
obtain {KN, ∆HN}, nonylphenol was injected into dried toluene.26 With respect to the friction losses, the reference test was the injection of toluene into toluene. The heat obtained (-2 µcal/ injection) was subtracted from the raw data prior to any further treatment. Once these parameters are fixed, the fitting of the model to the experimental data is done by means of the enthalpy of NP-ASP interaction (∆H) and the equilibrium constant (K). For the second model (model 2), a much simpler approach was applied. The heat developed in the nonylphenol reference data was subtracted from the raw data. The remaining heat was assumed to be due to the interaction ASP-NP and also to the effect of toluene on the aggregation state of asphaltenes. Asphaltenes are believed to associate stepwise,16 and in the past the self-association of asphaltenes has been successfully modeled with polymerization-type reactions29
Pn + P1 T Pn+1
[Pn+1] ) Kn+1[P1][Pn]
(5)
The equilibrium constants and enthalpies were considered to be the same for all the reactions
Figure 1. LM1 10 g/L asphaltenes and nonylphenol at several concentrations.
K ) K2 ) K3 ) ... ) Kn+1
(6)
∆Ha ) ∆Ha2 ) ∆Ha3 ) ... ) ∆Han+1
(7)
with eq 16. This model has three fitting parameters that are obtained in the optimization routine: the equilibrium constant K, the enthalpy ∆H, and the number of sites n. V is the volume of the cell.
In the asphaltene-resin experiments, asphaltenes are again modeled with polymerization reactions, while the ligands are considered to act as terminators of the polymerization type reaction.
Pn + T T PnT
[PnT] ) KTn[T][Pn]
(8)
To simplify the approach, it is considered that the equilibrium constants are the same as those of the propagation reactions, but the NP-ASP interaction is modeled with a different value of ∆H. The values of ∆Ha are calculated by applying eqs 5-7 to the titration of asphaltenes into dried toluene26 and the ∆H of the interaction asphaltene-ligand is used as a fitting parameter, together with the equilibrium constant K.
K ) K2 ) K3 ) ... ) Kn+1 ) KT1 ) KT2 ) ... ) KTn
(9)
∆Ha ) ∆Ha2 ) ∆Ha3 ) ... ) ∆Han+1
(10)
∆H ) ∆H1 ) ∆H2 ) ... ) ∆Hn
(11)
(13)
The heat developed in each injection is calculated by multiplying the enthalpy by the variation in amount of nonylphenol bound to asphaltenes ∆[RB]. The concentration of bound nonylphenol RB before and after each injection is calculated by means of eqs 14 and 15, where [A] is the total concentration of asphaltenes and [RT] and [R] are the total and free nonylphenol concentrations, respectively.
nK[R] [RB] ) [A] [R] + 1
(14)
[RT] ) [R] + [RB]
(15)
Results and Discussion
This terminator-propagator approach has been applied to model VPO measurements of asphaltenes in toluene solutions.30 Herein, asphaltenes are considered to contain only propagators, while nonylphenol is considered as a terminator. The calculated heat q is
q ) nA-N (mol)*(∆H) + nA-A (mol)*(-∆HA)
q ) ∆HA-RV∆[RB]
(12)
The last model, One Set of Independent Sites (ONE), has been widely applied in biochemistry.13 It considers that asphaltenes contain a number of sites available for the interaction, and they all have the same affinity for nonylphenol. Besides, the binding of one molecule is not affected by the neighboring sites. This means that two sites act as if there were very far from each other, even if they may be in the same molecule, which is a good approximation to large hydrocarbon structures such as asphaltenes. It allows the calculation of the average number of interaction sites at the same time as the enthalpy of association. After the subtraction of the reference data, the resulting heat is assumed to be related only to the binding of nonylphenol with asphaltene. Inherent in this, it is assumed that the injection of 290.0 µL in a cell of 1.46 mL does not lead to a substantial heat of dissociation of asphaltene aggregates, in accord with MerinoGarcia et al.29 The heat developed in each injection is calculated (29) Merino-Garcia, D.; Murgich, J.; Andersen, S. I. Pet. Sci. Technol., in press. (30) Agrawala, M.; Yarranton, H. W. Ind. Eng. Chem. Res. 2001, 40, 4664.
Nonylphenol-Asphaltene Interaction. The acidbase interaction between asphaltenes and amphiphiles seems to be quite complex and involves different mechanisms. The heat signal curves (Figure 1) show that there are at least three processes ongoing in the cell. Each peak represents the heat developed in one injection as a function of time. The negative peaks (2) represent an exothermic process, which is believed to be the interaction between free sites in asphaltene aggregates and nonylphenol (NP). As the sites available for the interaction become saturated, this contribution decreases in magnitude as the titration goes on. Once the heat developed in the interaction NPasphaltene becomes small, there is another endothermic peak (3) that appears, which represents the breaking of the H bonds among nonylphenol molecules (see the titration of nonylphenol into dried toluene, Casp ) 0 g/L in Figure 2). This heat contribution is more significant at high nonylphenol concentrations. There is another endothermic peak (1) that appears in the first injections, which varies in magnitude among the different asphaltenes. The origin of this peak is not fully understood. This peak is not observed in the absence of nonylphenol. It cannot be assigned to the dilution of nonylphenol because the positive peaks disappear after a certain number of injections and then appear again. If these two sets of peaks corresponded to the same process, the peaks would be there in all injections. It had been speculated as well that it could be caused by the control system. The action of the control system could be too strong and create a secondorder response. Nevertheless, similar experiments with
1476
Langmuir, Vol. 20, No. 4, 2004
Merino-Garcia and Andersen
Figure 2. Injection of 5 g/L of nonylphenol into Alaska 95 asphaltenes. Figure 4. Titration of 20 g/L of nonylphenol into 10 g/L asphaltenes.
Figure 3. Injection of 20 g/L of nonylphenol into Alaska 95 asphaltenes.
other substances gave peaks of the same magnitude or even higher and this second-order response was not observed. Since the nonylphenol-asphaltene interaction has been shown to be exothermic, it is assumed to represent the dissociation of asphaltene-asphaltene bonds due to the presence of nonylphenol. The work cited in the Introduction that reports the incapacity of nonylphenol to disperse asphaltenes is not in disagreement with this statement. That work8 was performed on asphaltene colloids in n-heptane, while the present experiments are in toluene, where asphaltenes only form aggregates of four to six molecules that cannot be considered colloids. Nonylphenol may have enough dispersing power to substitute an asphaltene molecule in an interaction site. The nonylphenol-asphaltene interaction is more evident at high asphaltene concentration (Figures 2 and 3). At low asphaltene concentrations, the peaks are positive, which means that the main process in the cell is the dissociation of nonylphenol associates. The amount of asphaltenes present is very low, and the peak of the interaction nonylphenol-asphaltenes is masked by the big contribution of the breaking of NP-NP bonds. As the concentration of asphaltenes increases, the negative peak of the interaction ASP-NP starts to grow and at high asphaltene concentrations it masks the endothermic breaking of NP-NP bonds. As expected, the heat developed with 20 g/L of nonylphenol is greater than that with only 5 g/L, as there are more sites available for hydrogen bonding. The integration of the peaks gives the heat developed in each injection. It is important to notice that the heat developed is very low, in the order of 10-100 µcal/injection.
Figure 5. Fit of model 1 to the titration of 10 g/L Alaska 95 Asp with 5 and 40 g/L of NP. The dots represent the experimental data, and the lines represent the model.
This high sensitivity and low detection limit are only available in the last generation of calorimeters. ITC experiments with Cs ) 20 g/L nonylphenol show that the heat developed is different for each asphaltene (Figure 4). LM1 and Alaska 95 show an abrupt change in slope at around 1 g/L. This may be related to the end of the dispersive effect of NP on asphaltene aggregates. This process seems to have a faster decay in heat developed than the rest of processes, and it is quickly masked by the exothermic peaks. There are several factors that may have an influence on the heat developed, but it is assumed that the one with the highest heat developed is the one with most interaction with nonylphenol. The effect of the dilution with toluene on the aggregation state of asphaltenes is also relevant, but the volume injected is small and the heat developed in this process would be smaller than the interaction NP-ASP. This suggests that Alaska 95 is the asphaltene with greatest acid-base capacity of the ones investigated in this study. Modeling. The application of model 1 to the titration of asphaltenes with nonylphenol is not very satisfactory (Figure 5). The figures have the x axis in reverse order due to the fact that the concentration of asphaltenes decreases as the titration goes on. The optimal {∆H, K} are shown (Table 1), together with the values of the sum of squares (SS). Some of the experiments present good fits, but some have too large values of SS. Tests with 40 g/L NP give much larger SS values than tests with lower NP concentrations. Besides, the values of ∆H vary a lot
Interaction of Asphaltenes with Nonylphenol
Langmuir, Vol. 20, No. 4, 2004 1477
Table 1. Results of Model 1 with Asphaltenesa CNP (g/L)
LM1 Asp. Ki SS
∆H
KU Asp. Ki SS
∆H
A95 Asp. Ki SS
Casp cell ) 10 g/L -3.5 400 -12.3 -3.9 300 0.8 -9.2 125 -1.4 -1.7 450 73.2 -5.9 125 3.1 -3.1 125 -0.7 -7.0 175 211.0 -5.6 175 182
5 20 40
-1.0 700
5 20 a
∆H
Casp cell ) 1 g/L 5.8 -0.9 700 4.3 -8.0 250 -10.7 275 122.0
-1.5
∆H in kJ/mol, K in L/mol, and SS in µcal/injection. Table 2. Results of Model 2 with Asphaltenesa
LM1 Asp. C NP (g/L) ∆Hi Ki SS 5 20 40 100
KU Asp. Ki SS
∆Hi
A95 Asp. Ki SS
C asp cell ) 10 g/L -5.4 2173.5 -9.4 -3.5 3702.9 -0.4 -5.3 1234.7 -0.3 -5.2 304.6 -1.2 -3.7 172.8 -0.9 -5.6 135.0 -0.5 -6.4 101.1 -15.9 -14.2 -5.9 100.8 -42.9
5 -2.4 20 a
∆Hi
284.4
C asp cell ) 1 g/L -0.5 -1.8 234.5 -1.0 -4.8 -1.8 154.2 -0.7 -5.0
127.1 75.3
Figure 7. Fit of model 2 to LM1 10 g/L asphaltenes titrated with several nonylphenol concentrations: (O) 5 g/L NP; (+) 20 g/L NP; (0) 40 g/L NP; (×) 100 g/L NP and 3 µL per injection.
-0.3 -0.5
∆H in kJ/mol, K in L/mol, and SS in µcal/injection.
Figure 8. Fit of model 2 to the titration of 10 g/L Alaska 95 Asp with 20 g/L NP.
Figure 6. Fit of model 2 to the titration of 10 g/L Alaska 95 Asp with 5 and 40 g/L of NP.
from one test to another. Model 2 gives values of ∆H that are more similar among the tests with different NP concentrations (Table 2). The goodness of fit is remarkably good for most of the experiments (Figure 6). Nevertheless, the equilibrium constants K are still concentration dependent. They tend to decrease as the concentration of nonylphenol increases. This could be interpreted in terms of affinity: at high nonylphenol concentration, it would have more tendency to self-associate than to interact with asphaltenes. At low concentrations of NP, the proportion of monomeric nonylphenol is higher, allowing a higher interaction with asphaltenes that implies a higher value of the equilibrium constant K. Buenrostro-Gonzalez et al.21 applied the SAFT-VR equation to model the precipitation of Maya asphaltenes from crude with n-alkanes, obtaining an enthalpy of interaction of -3.3 kJ/mol, as a fitted parameter. The agreement between these data and those in Table 5 is very satisfactory. It is observed that there is less heat of interaction NPASP per gram of nonylphenol injected as the concentration of NP increases (Figure 7). This supports the above explanation of why K decreases following the same trend.
It must be remembered that this approach is probably an oversimplification of the system. The entire asphaltene fraction is represented by only one molecule. The polydispersity in both polarity and aromaticity2 of the asphaltene fraction is simply not taken into account. Some of the asphaltene molecules would have more affinity than others to nonylphenol, and some would just not be active in these titrations.29 Besides, some of the sites that could be used by nonylphenol may be involved in asphalteneasphaltene bonds. The average molecular weight is set arbitrarily to 1000 g/mol. Again, the polydispersity in molecular size is not considered in this model. These simplifications are relevant, but at this stage of the investigation there is no choice. Once the picture of asphaltene self-association is better defined, it would be possible to come up with a more detailed model without using too many fitting parameters. Still, the values of ∆H is well within the range expected for hydrogen bonding showing that the modeling has indeed some merits in understanding the process. Due to these simplifications, the model is not able to fit the data when process 1 (see Figure 1) is observed. This is probably due to the complexity of the signal. The calorimeter measures the global heat, while model 2 only considers nonylphenol-asphaltene association. Once this process 2 is over, model 2 can fit very well the data because the only heat measured by the calorimeter is developed by a nonylphenol-asphaltene reaction (Figure 8). Interaction Sites. The IR spectrum of nonylphenol in toluene has been obtained in the -OH stretching region
1478
Langmuir, Vol. 20, No. 4, 2004
Merino-Garcia and Andersen
Table 3. Capacities in mol of NP/kg of ASP by IR Spectroscopy LM1 ASP
NP ) 5 g/L
NP ) 10 g/L
Alaska 95
NP ) 1.83 g/L
NP ) 18.3 g/L
C ) 1 g/L C ) 10 g/L
7.1 0.9
13.6 1.1
C ) 1 g/L C ) 10 g/L
1.04 0.16
3.19
Figure 9. Influence of Alaska 95 asphaltenes on the IR spectra of 20 g/L NP.
(3600-3300 cm-1). The band has been deconvoluted into four peaks located at 3560, 3550, 3462, and 3390 cm-1. Each of these peaks is associated to a state of the -OH group in the solution (free, bonded by the O atom, bonded by the H atom and bonded by both H and O atoms, respectively). The location of the peaks is different from that found in the literature for CCl4 solutions of phenol.31 The solvation effect of toluene is very strong and shiftes the peaks to lower wavenumbers. IR spectroscopy of nonylphenol shows that the band assigned to free -OH (3560 cm-1) decreases with the presence of 10 g/L Alaska 95 asphaltenes (Figure 9). The decrease is assigned to the amount of nonylphenol attached to asphaltenes, as reported previously for the interaction with phenol by means of the phenol interaction value.32-33 It is considered that the presence of asphaltenes does not modify the equilibrium between free and self-bonded nonylphenol molecules.12 The capacity of interaction with nonylphenol of Alaska 95 is estimated to be 3.19 mol of NP/kg of Asp at that particular concentration. Chang and Fogler12 found that the hydrogen bonding capacity of asphaltenes with dodecylphenol was 1.6-2.0 mol/kg, following the same procedure. This unit has been chosen because the use of molar basis depends on the molecular weight assumed. Besides, an average Mw of 1000 units is often reported as a good estimation, so the molar number of sites per kilogram is also a reliable approximation to the average number of sites per asphaltene molecule. The amount of NP attached to asphaltenes varies significantly with the concentration of both species (Table 3). As expected, the amount of NP attached increases with increasing NP concentration, but it decreases as the concentration of asphaltenes rises. Some of the potential sites for NPASP interaction are probably occupied in asphalteneasphaltene bonds at high asphaltene concentration, as suggested in studies of the interaction phenol-asphaltenes.34 To calculate the maximum capacity of interaction with asphaltenes, a high concentration of nonylphenol (300 or 468 mM) was injected into a low concentration of asphaltenes (1 g/L). This dilution ensures that asphaltenes have a low aggregation state and most of the sites are available for the interaction with nonylphenol. The (31) Vinogradov, S. N.; Linell, R. H. Hydrogen bonding; Van Nostrand Reinhold Company: New York, 1971. (32) Barbour, R. V.; Petersen, J. C. Anal. Chem. 1974, 46, 273. (33) Andersen, S. I. Fuel Sci. Technol. Int. 1994, 12 (11&12), 1551. (34) Siddiqui, M. N. Pet. Sci. Technol. 2003, 21 (9&10), 1601.
Figure 10. Titration of 100 g/L of NP into dried toluene and 1 g/L LM1 ASP (volume per injection ) 3 µL).
Figure 11. Calculation of maximum number of sites in LM2 asphaltenes.
exothermic interaction NP-ASP is rapidly compensated by the high heat developed in the break of H bonds among nonylphenol molecules (Figure 10). The subtraction of the reference data (injection of NP into dried toluene) gives the heat developed in the interaction NP-ASP. This heat reaches a value of zero when all sites have been saturated. The concentration of nonylphenol (C*) at which the sites become saturated has been calculated by drawing the trendline of the points in the linear region of the curve (Figure 11). If it is assumed that all nonylphenol molecules in the cell would interact with asphaltenes, C* gives the number of sites n available for interaction. n varies from 6 to 8 depending on the asphaltene. In reality, not all nonylphenol molecules are attached to asphaltenes, so these values can be considered an upper limit. Model 2 was first used to obtain the thermodynamic parameters of NP-asphaltene interaction (Table 4). ∆H is significantly lower than in the experiments previously
Interaction of Asphaltenes with Nonylphenol
Langmuir, Vol. 20, No. 4, 2004 1479
Table 4. Results of the Fitting of Model 2 to Experiments with High Nonylphenol Concentration (300 mM)a n ∆H (kJ/mol) K ∆G (kJ/mol) ∆S (J/mol) a
LM1
Lagrave
Ca30
NM1
Yagual
A95
KU
LM2
1 -5.3 89.1 19.8 -11.3
1 -1.5 168.1 37.7 -12.9
1 -0.9 225.2 42.1 -13.6
1 -1.6 438.2 45.3 -15.3
1 -1.2 545.9 48.4 -15.9
1 -1.9 254.2 39.8 -14.0
1 -1.0 591.3 49.8 -16.1
1 -1.8 207.2 38.4 -13.4
KU
LM2
LM1 was performed with 468 mM nonylphenol. Table 5. Results of the Fitting of ONE Modela LM1
Lagrave
Ca30
NM1
Yagual
Alaska 95
n 2.6 ( 0.9 4.6 ( 0.9 5.1 ( 0.7 5.0 ( 0.6 4.1 ( 0.8 4.3 ( 0.4 4.8 ( 0.6 3.2 ( 0.7 ∆H (kJ/mol) -2.0 ( 0.7 -0.07 ( 0.02 -0.10 ( 0.02 -0.06 ( 0.01 -0.06 ( 0.02 -0.19 ( 0.02 -0.07 ( 0.01 -0.3 ( 0.1 ∆G (kJ/mol) -11.1 -17.1 -15.7 -18.8 -17.4 -16.0 -18.5 -14.4 ∆S (J/mol) 29.9 56.2 51.4 61.9 57.3 52.1 60.8 46.3 K 81 ( 6 900 ( 500 500 ( 200 2000 ( 1000 1000 ( 600 600 ( 100 1600 ( 900 300 ( 60 a
n in sites per molecule (Mw ) 1000 g/L). Table 6. Heteroatom Content from Elemental Analysisa mol of O/mol of ASP mol of N/mol of ASP mol of S/mol of ASP a
KU
LM1
NM1
A95
CA30
LM2
3 1.2 2.1 6.3
1.4 1.3 1.5 4.2
1.4 1.3 0.9 3.6
1.2 0.9 1 3.1
1.1 0.7 2.4 4.2
2 1.2 1.2 4.4
Mw assumed ) 1000 g/mol.
Figure 12. Fit of ONE model to nonylphenol-asphaltene experiments.
shown (see Table 2); on the contrary, the equilibrium constants are greater. ∆G is negative, leading to the conclusion that the process is spontaneous. The values obtained are inside the interval assigned to hydrogen bonding (between -10 and -40 kJ/mol).35 The contribution of ∆S to the free energy is greater than the one of ∆H: this may indicate that the process is entropically driven. Model 2 has an important drawback due to the fact that the number of sites per asphaltene molecule is fixed to 1 (see eq 11). This limits the predictive capacity of the model and creates a greater dependence of K and ∆H with the Mw assumed. ONE model (see eqs 16-18) was proposed, as it includes the number of sites as a third fitting parameter. It is as well able to fit successfully the experimental data (Figure 12). Table 5 compiles the thermodynamic parameters of the interaction. ∆H is exothermic, as seen with the previous model, but the values are extremely low. They are 1 order of magnitude lower than the ones obtained with model 2. This is considered a deficiency of the model. The greater equilibrium constants make ∆G and ∆S be similar to the previously discussed. The number of sites ranges from 2.6 to 5.1 sites per kilogram; the agreement between these values and the IR data is as well satisfactory. Elemental analysis (Table 6) shows that the amount of heteroatoms per kilogram of asphaltenes is also in that range. This may indicate that the hydroxyl group of nonylphenol is mainly attached to asphaltenes through these heteroatoms. Nevertheless, the capacity of nonylphenol to form
bonds with the π-orbitals of the aromatic rings cannot be neglected. For instance, the OH-π intramolecular bond in 2-benzylphenol is reported to have an enthalpy of formation of -1.38 kJ/mol.36 This enthalpy is in the same range as the ones reported here, leading to the conclusion that the upper limit of six to eight sites of interaction found by means of C* is not such a bad estimation, if both heteroatoms and aromatic rings are considered to form bonds with nonylphenol. Experiments were as well performed with greater nonylphenol concentration (468 mM). These experiments presented a larger scattering of the data (Figure 13) due to the fact that very small volumes were injected (2 µL). However, the data were useful to check the parameters obtained with 300 mM nonylphenol. Figure 13 shows that these parameters are as well able to give a good performance in experiments with a greater nonylphenol concentration. The average number of sites (n) obtained are in the same range as those reported or used by other researchers. Leon et al.37 found 6.7 molecules of NP per asphaltene molecule asphaltene in adsorption studies on particles in
(35) Joesten, M. D.; Schaad, L. J. Hydrogen Bonding; Marcel Dekker: New York, 1974.
(36) Bellamy L. J. Advances in Infrared Group Frequencies; Methuen & Co. Ltd.: London, 1968; p 246.
Figure 13. LM2 experiments with nonylphenol: (O) Vinjected ) 3 µL; C syringe ) 300 mM nonylphenol; (×) Vinjected ) 2 µL; C syringe ) 468 mM; (s) fit of ONE model.
1480
Langmuir, Vol. 20, No. 4, 2004
Merino-Garcia and Andersen Table 7. Results of the Application of Model 2 to Asphaltene Fractionsa crude
CASP (g/L)
CNP (g/L)
∆H (kJ/mol)
K (L/mol)
SS
KU SOL
1 1 10 10 1 10 10 1 10 10 1 10 10
5.0 20.0 5.0 20.0 5.0 5.0 20.0 5.0 5.0 20.0 5.0 5.0 20.0
-2.1 -4.1 -3.2 -3.7 -2.1 -2.1 -1.8 -4.9 -4.9 -6.4 -3.4 -3.2 -5.0
141.9 63.8 1961.5 223.6 170.1 5258.6 718.4 143.9 1081.5 101.0 115.7 1267.2 95.1
0.4 0.8 0.4 3.5 1.8 0.4 1.4 2.0 0.1 0.6 0.7 0.2 2.9
KU INS LM1 SOL LM1 INS
a
Figure 14. Titration of 10 g/L of KU fractions with 20 g/L NP and fit of model 2. The first chart shows the raw data and the second the integrated heats.
n-heptane. Murgich et al.38 used nine sites available for resin molecular simulations in an asphaltene aggregate, based on the most probable interaction sites in a model molecular structure from Athabasca sand asphaltene. Wu et al.39 used six sites of interaction per asphaltene molecule in their SAFT calculations. Buenrostro-Gonzalez et al.21 used three to four sites to successfully fit onset precipitation data with the SAFT equation. However, these numbers are only an estimation of n, as some of the potential sites for NP-Asp interaction are actually occupied in asphaltene-asphaltene bonds. Both ONE and model 2 are equally able to fit the experimental data. Model 2 catches better the complexity of self-association, while the ONE model only considers one reaction. However, model 2 is not able to give the number of sites, as this parameter is fixed to 1. The ∆H values obtained with model 2 are in the range expected, while the ONE model gives surprisingly low values of enthalpy. Both are just a first attempt to model asphaltene-nonylphenol interaction. It is observed that the equilibrium constants vary quite arbitrarily from one experiment to another. The models are unable to catch the complexity of the asphaltene mixture. This is again due to the simplicity of the approach. The assumptions made in both cases are very strong, as they apply a monodispersed approach to a highly polydispersed system. Still, these simple models are able to catch the behavior of nonylphenol and asphaltenes in toluene solutions. Asphaltene Fractionation. The titration of KU (Figure 14) and LM1 fractions with nonylphenol shows that there is a much lower heat developed when the INS fraction is in the cell, and the heat reaches faster a constant value. The fact that the soluble fraction has a greater heat of interaction with nonylphenol confirms that the SOL fraction is more polar than the INS fraction.29 The application of model 2 to the fraction is as satisfactory as that for the entire asphaltenes (Table 7). ∆H in kJ/mol of INS is significantly smaller of that of SOL: if it assumed that the acid-base interaction is the same for all fractions, ∆H should also be the same. This (37) Leon, O.; Rogel, E.; Urbina, A.; Andujar, A.; Lucas, A. Langmuir 1999, 15, 7653. (38) Murgich, J.; Abanero, J. A.; Strausz, O. P. Energy Fuels 1999, 13, 278. (39) Wu, J.; Prausnitz, J. M.; Firoozabadi, A. AIChE J. 1998, 44 (5), 1188.
∆H in kJ/mol, K in L/mol, and SS is in µcal/injection.
implies there are fewer bonds formed per mole of asphaltenes; therefore, INS would be less active in the acidbase interaction. The problem is the equilibrium constant of INS fraction, which becomes larger than the one for SOL, which implies there are more moles of PiT formed in the INS fraction. This goes against what was expected intuitively: it seems reasonable that SOL has more bonds formed, as the exothermic heat developed is greater. This is considered again a deficiency of the simple fitting model. Conclusions The results show that the calorimetric experiments open a new way of investigating association in petroleum, but so far, the analysis is still hindered by the lack of knowledge about asphaltenes. The numerous assumptions made to calculate the heats of interaction between nonylphenol and asphaltenes suggest that the numerical results should be accepted with caution. In light of the very complex mixtures analyzed, the approach used is very simplistic, as each process is characterized with one set of values of ∆H and K. A more detailed model would be necessary, but that would greatly increase the number of parameters to be regressed. Besides, a better understanding of asphaltene association is needed to develop a more consistent model. Still, some interesting results have been obtained: experimental values of ∆H have been obtained, with the idea of extending this study to native resins, for which there are no data available in the literature. The specific number of sites has been obtained, and there is good agreement with the number of heteroatoms in asphaltenes and also with the usual estimations in asphaltene modeling. The simple models used present serious problems related to the complexity of the mixtures they have to handle. In any case, the maximum number of sites has been obtained directly from the titration curves by means of the calculation of C*. These experimental values could be very useful to reduce the number of fitting parameters in the development of asphaltene models. Besides, it was proved that the fractionation with a mixture of toluene and acetone gives a soluble fraction that is more polar than the insoluble fraction, confirming the highly polydispersed nature of asphaltenes. Acknowledgment. The authors acknowledge the Danish Technical Research Council (STVF under the Talent Program) for financial support and Mr. Zacarias Tecle for his great help in the laboratory. LA035875G